the medication of sadness · meds for 4 mos., placebo for 12 mos. meds for 16 mos. ct for 4 months,...
Post on 30-Mar-2021
3 Views
Preview:
TRANSCRIPT
The Medication of Sadness
Robert J. DeRubeis
Samuel H. Preston Term Professor
in the Social Sciences
Professor and Chair of Psychology
School of Arts and Sciences
University of Pennsylvania
Center for Neuroscience and Society
June, 2013
Depression, a Disorder of Mind and Brain
• Lifetime prevalence of Major Depressive Disorder:
– 10% in men
– 20% in women
• Estimated annual workplace losses in U.S.: $50 billion
• Substantial contributors to incidence and onset:
– stressful life events (losses)
– heritable genetic risk
– interaction of genes X environment (e.g., promoter region of the
serotonin transporter gene X life events)
• Genetic and environmental effects each impair the structure
and function of:
– the primitive (limbic) brain – supports emotions, basic drives
– the neocortex (frontal lobes) – supports executive functions,
including those that regulate emotion
Profound depression, from the inside
“I am now the most miserable man living.
If what I feel were equally distributed
to the whole human family, there would
not be one cheerful face on earth.”
Profound depression, from the inside
“I am now the most miserable man living.
If what I feel were equally distributed
to the whole human family, there would
not be one cheerful face on earth.”
Abraham Lincoln
Evidenced-based Somatic Treatments
• Electroconvulsive Therapy
• Most effective, but least widely used
• Cost, aversion to the procedure, memory effects explain limited use
• Antidepressant medications (ADM)
• Introduced in 1950s (MAO-I; e.g., Nardil, selegiline patch)
• 1960s: tricyclics (TCA; e.g., imipramine, Elavil)
• 1980s: SSRIs (Prozac, Paxil)
• 2000s: NSRIs (Effexor)
• Principle clinical differences between ADM classes:
• Tolerability; nature and severity of side effects
• Suicide risk potential
• All produce, on average, 15%-25% increment in response rates,
relative to pill-placebo (e.g., 60% vs. 40%), in patient populations on
which they have generally been tested
Evidenced-based Psychological Treatments for Major Depression
• Cognitive Therapy (CT)
• Most extensively researched, most widely practiced
• Compares favorably with ADM, even in more severe cases
• Only treatment with substantial evidence of relapse prevention
• Interpersonal Therapy (IPT)
• Has fared well as acute treatment in clinical trials, but with slower onset
of effect than ADM or CT
• Effective as a maintenance treatment, but not as prevention
• Readily accepted by clinicians who trained psychodynamically
• Behavioral Activation (BA)
• Excellent results in two clinical trials, both conducted in Seattle
• May be better-suited for dissemination than CT or IPT
Key findings from the National Health and
Nutrition Examination Surveys, 2005–2008
Americans 12+ yrs. old: 11% are on ADMs
– Among those with no symptoms: 8%
– Mild symptoms: 19%
– Moderate symptoms: 28%
– Severe symptoms: 33%
Key findings from the National Health and
Nutrition Examination Surveys, 2005–2008
Of all those on ADMs
– 60% (7% of pop.) have been on them for 2+ years
– 14% (1.5% of pop.) have been on them 10+ years
– Less than 1/3rd of those on 1 ADM – and less than ½ of
those on 2+ ADMs – have seen a mental health
professional in the past year
Antidepressant use among undiagnosed:
Patient’s complaint
• Premenstrual tension
• Sleep disturbances
• Migraines
• Feeling tired
(Mojtabai & Olfson, 2011)
National Trends in Outpatient Treatment of
Depression (Olfson et al 2002)
0%
10%
20%
30%
40%
50%
60%
70%
80%
1987 1997
Antidepressants
Psychotherapy
General Medical
Olfson et al 2002 (JAMA) % Antidepressants increased (p<.001)
% Psychotherapy decreased (p = .006)
% General medical visits unchanged (ns)
Reasons for Trends over Time (Olfson et al 2002 JAMA)
• Safety and ease of prescribing newer antidepressant medications (SSRIs), relative to older medications
• Federal government (also NAMI, etc.) embarked on public health campaign to educate public about depression
• Promotion of sales through vigorous advertising campaigns to primary care doctors, and to consumers
• Growth of managed care resulted in shifts from specialty to primary care medical management
– Psychotherapy reimbursed less generously than medications
– Primary care physicians more likely to use medications
Antidepressant prescriptions in the U.S.
• 80% of prescriptions are made by physicians who
are not psychiatrists
– 62% written by PCPs
• Antidepressants increasingly prescribed for
undiagnosed conditions (55-70%)
(Cascade et al., 2008; Mojtabai & Olfson, 2011)
Primary Care Physicians (PCPs)
• Treat at least 50% of depression cases
• 10-30% of patients in primary care are depressed
• > 50% of insured report they prefer going to a PCP
for mental health issues
(Mickus, Colenda, & Hogan, 2000 ; Simon et al., 2002; Wang et al., 2005)
Treated vs. untreated depression
in primary care
• After 1 year unrecognized cases did significantly
better than those recognized (Goldberg et al., 1998)
• After 1 year patients not given drugs did significantly
better than those receiving drugs (Goldberg et al.,
1998)
• Unrecognized cases had comparable outcome after
12/39 months as recognized cases (Kamphuis et al.,
2011)
Americans’ Attitudes Toward
Psychiatric Medications (1998–2006)
• 68% think medications help people feel better about
themselves (vs. 60% in ’98)
• 83% think medications help deal with day-to-day
stresses (vs. 78% in ’98)
• 47% would take medications to cope with life stresses
(vs. 36% in ’98)
(Mojtabai, 2009)
Reasons for Reduction in use of
Psychotherapy over Time (DeRubeis)
• Lack of public education about the effectiveness of psychotherapies
• Lack of availability of professionals trained in evidence-based psychotherapy practice
• Persistence of the belief that psychotherapy is more expensive than medications (many reasons for this)
• Consumers’ demand for antidepressants
– high expectations
– low cost (co-pays)
– knowledge that they take little time from busy schedules and do not require speaking about embarrassing topics
Have the shifts in treatment practices
yielded greater benefits? Decreased costs?
• Rates of depression, and chronicity of depression, continue to increase
• Relapse rates on the most common treatments are as high as 50-80%
• Lifelong treatment is recommended for patients with recurrent depression (high cost)
• Patients often discontinue their medicines AMA; “discontinuation syndrome” can be quite unpleasant
• Question: Why haven’t we developed more effective -- or more efficient -- treatments and treatment strategies?
Focus of Clinical Trials
• Depressed patients with high severity symptoms
– Group of greatest concern
– Most likely to reveal difference between active
treatment vs. control treatment (e.g., placebo)
Post-treatment Symptom Severity (Hamilton Scale) for
Patients Who Were Severely Depressed Prior to Treatment
11.7 11.9
0
2
4
6
8
10
12
14
16
18
Elkin*
(N=53)
Rush*
(N=26)
Murphy*
(N=22)
Hollon*
(N=68)
DeRubeis,
Hollon, et al.
(n=180)
Pooled
(n=349)
De
pre
ss
ion
Se
ve
rity
AntidepressantsCognitive Therapy
*(From DeRubeis, Gelfand, Tang, & Simons, 1999) ;
Time to Relapse (n = 35 per group)
24%
53%
69%
0
25
50
75
100
0 4 8 12
Months in Continuation
Per
ecen
t N
ot
Rel
ap
sin
g
Meds/Placebo Meds+Meds Cog. Ther.
14%
30%
40%
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
Meds for 4 mos.,
Placebo for 12 mos.
Meds for 16 mos. CT for 4 months, 3
sessions in next 12
mos.
Sustained Improvement through 16 months
0
500
1000
1500
2000
2500
3000
3500
4000
4500
1 3 5 7 9 11 13 15 17 19 21 23 25 27
Months in Treatment
Co
st
in D
olla
rs
Meds
Therapy
Cumulative Direct Costs of Meds vs. Therapy
How did we get here?
• 1950s – excitement about the prospect of understanding and
treating severe (real) depression by identifying and altering
crucial (simple) brain mechanisms
• Destigmatization, fostered by advocacy groups, largely
funded by the pharmaceutical industry
• Expansion of the definition of depression (DSM-III)
• Introduction of Prozac – and other SSRIs – whose side
effects could be insidious, but not scary or unsettling
• Promotion of ADMs by the most powerful and persuasive
members of the psychiatric establishment
How did we get here?
• The experiences of ADM-takers:
– Worked on their depression
– Other positive effects (calming, etc.)
– Placebo effects
– Discontinuation effects (confirming the benefits of the ADMs)
• Major campaigns to increase awareness of the benefits
of ADMs (TV ads, websites, textbooks, etc.)
What happened next?
• Treatment of depression moved more and more to
primary care (where ADMs are readily prescribed)
• Patient demand for medications increased; awareness of
other means of dealing with depression did not
Major depressive disorder vs.
understandable sadness
Major depressive disorder vs.
understandable sadness
Allan Horwitz &
Jerome Wakefield
Sadness
• Universal emotion
– Cross-cultural
– Present in animals
• Functional loss response
– Interpersonal
– Hierarchical/social structure
– Failure to achieve goals
• Has biological correlates
• Remits with recovery of loss or passage of time
Major Depressive Disorder
• Long recognized as an affliction
– Cross-cultural
– Animal models
• Often (85%) occurs in reaction to a stressor,
but seen as dysfunctional
• Has biological correlates
• Can “spontaneously” remit
Features of a mental disorder
A. Clinically significant psychological pattern occurring in
an individual
B. Associated with distress, disability, or risk of suffering
death, pain, disability, or loss of freedom
C. Must not be merely an expectable and culturally
sanctioned response to a particular event
D. A manifestation of a behavioral, psychological, or
biological dysfunction in the individual
E. Deviant behavior is NOT a mental disorders unless as
a symptom of a dysfunction in the individual
According to the DSM IV TR
• “Periods of sadness are inherent aspects of human
experience. These should not be diagnosed as MDD
unless criteria are met for”:
– Severity (5 or more symptoms)
– Duration (≥2 weeks)
– Distress or impairment
(APA, 2004; p. 355)
According to the DSM IV TR
• “Periods of sadness are inherent aspects of human
experience. These should not be diagnosed as MDD
unless criteria are met for”:
– Severity (5 or more symptoms)
– Duration (≥2 weeks)
– Distress or impairment (Wakefield, Schmitz, & Baer, 2010)
(APA, 2004; p. 355)
Foreword to the Loss of Sadness
“…the DSM is not consistent even in applying
its own definition of mental disorder to the
diagnostic criteria sets for specific disorders.”
-Robert Spitzer
Loss of Sadness
• Sadness is an adaptive loss response
• Major Depression should involve a dysfunction of normal loss responses
• The symptoms of MDD can occur in response to a life stressor, without being indicative of dysfunction
• By ignoring the context in which symptoms occur, the DSM confuses sadness with MDD
Case
A 64-year-old married man has developed feelings of
sadness and emptiness, lack of pleasure in activities,
insomnia, fatigue and lack of energy, and feelings of
worthlessness.
He is not interested in seeing friends and seems unable to
concentrate on anything.
He yells at his wife when she attempts to console him and
rejects her efforts to comfort him.
Case
The feelings were triggered 2 weeks before when the company the
man worked for unexpectedly fired him as part of a corporate
downsizing, just 6 months before he would have qualified for the
company's retirement plan.
One of the major reasons the man chose to work for the company
and then spend two decades with it had been the prospect of
generous retirement benefits.
The loss of these benefits means that he and his wife will have
very little retirement income other than Social Security.
Case
Subsequently, the couple is forced to sell their house and
move to a small apartment.
The man finds part-time work that, along with Social
Security, provides barely enough resources to sustain him
and his wife.
He remains bitter about how he was treated, but his
symptoms gradually subside over time.
Proposal, from “Loss of Sadness”
Sadness
1. Context specific to losses
AND
2. Proportional in intensity and
duration
1. Cognitive
2. Affective
AND
3. Wanes with time, changes in
circumstances, and internal
coping
Proposal, from “Loss of Sadness”
Sadness
1. Context specific to losses
AND
2. Proportional in intensity and
duration
1. Cognitive
2. Affective
AND
3. Wanes with time, changes in
circumstances, and internal
coping
Major Depressive Disorder
Proposal, from “Loss of Sadness”
Sadness
1. Context specific to losses
AND
2. Proportional in intensity and
duration
1. Cognitive
2. Affective
AND
3. Wanes with time, changes in
circumstances, and internal
coping
Major Depressive Disorder
1. Occurs “out of the blue”
OR
2. Disproportionate in intensity
and duration
1. Cognitive
2. Affective
OR
3. Persists despite changes in
circumstances or passage of
time
Loss of Sadness:
Endogenous vs. Reactive Revisited?
• Different assumption about the role of life
events
• No assumption about symptom pattern
• No assumption about treatment response
Why does it matter?
• DSM as an obstacle to neuroscience (Nestler & Hyman,
2011)
• Research on etiology of mental disorder
• Unnecessary treatment
• Misleading prevalence rates
Bereavement (Kendler et al., 2008)
• Individuals with bereavement-related MDD and
MDD related to other stressors are comparable in:
– Demographics
– Personality traits
– Symptoms of depression
– Co-morbidities
• The bereavement exclusion should be removed
from DSM
Bereavement (Wakefield et al., 2007)
• Individuals with bereavement-related MDD and
MDD related to other stressors are comparable
• Complicated vs. uncomplicated MDD differ in:
– Impairment
– Melancholic symptoms
– Duration of episode
• The bereavement exclusion should be expanded
in DSM
What do other people think?
• Lay people distinguish between depression
coming “out of the blue” vs. depression following
a life stressor (Holzinger et al., 2009)
• Psychologists extend the bereavement exclusion
to other stressful life events (Kim et al., 2012)
• Mental disorders must not be merely an
expectable and culturally sanctioned response to
a particular event
What Should We be Doing to Develop a
Rational Approach to Treatment?
• Develop methods to determine whether “treatment” is
appropriate, vs.:
• Watchful waiting
• Psychoeducation
• Coaching re diet, exercise, social engagement, etc.
• For those for whom treatment is indicated, ask more than
whether a treatment is “evidence-based” (yes vs. no)
The enterprise of clinical science should aim to
provide, for any treatment…
The enterprise of clinical science should aim to
provide, for any treatment…
• unbiased estimates
The enterprise of clinical science should aim to
provide, for any treatment…
• unbiased estimates of the magnitude…
The enterprise of clinical science should aim to
provide, for any treatment…
• unbiased estimates of the magnitude…
– of its costs, benefits, and ancillary effects…
The enterprise of clinical science should aim to
provide, for any treatment…
• unbiased estimates of the magnitude…
– of its costs, benefits, and ancillary effects…
• both short-term and long-term…
The enterprise of clinical science should aim to
provide, for any treatment…
• unbiased estimates of the magnitude…
– of its costs, benefits, and ancillary effects…
• both short-term and long-term…
– relative to alternatives (nothing, placebo, other treatments)…
The enterprise of clinical science should aim to
provide, for any treatment…
• unbiased estimates of the magnitude…
– of its costs, benefits, and ancillary effects…
• both short-term and long-term…
– relative to alternatives (nothing, placebo, other treatments)…
» across the range of those given the diagnosis…
The enterprise of clinical science should aim to
provide, for any treatment…
• unbiased estimates of the magnitude…
– of its costs, benefits, and ancillary effects…
• both short-term and long-term…
– relative to alternatives (nothing, placebo, other treatments)…
» across the range of those given the diagnosis…
» and for important subsets of those with the diagnosis
FDA approval requirements
• Two studies* in which the new drug beats placebo at the .05
level on the primary measure of depressive symptom severity
• Neither the number of trials it took to get to two -- nor any
estimate of the magnitude of the effect -- plays a discernible
role in the FDA verdict
• Verdict is “thumbs up” or “thumbs down” (not an estimate of
degree or likelihood of benefit)
• No consideration of long-term benefit or harm, despite
knowledge that depression is or can be chronic, and
concerns about long-term side effects
Research Design Elements in FDA Registration Trials that
Should Maximize the Likelihood of Obtaining an Effect
– Include only patients who are in the “golden zone”
– Employ a placebo run-in, to exclude placebo
responders
– Do not use “active” placebos
Even so…
Overview of Turner study N Engl J Med. 2008 Jan 17;358(3):252-60
• FDA reviews
• 12 antidepressants starting with Prozac
• Cohort of 74 trials registered with FDA
• Track each study into published literature
• Two questions: – Was the study published?
– If published, how did the published results compare with the FDA results?
Breakdown by drug -- Journal view
Breakdown by drug -- FDA view
FDA decision Publication fate
Spun trials
Effect size (ES) with CIs:
FDA vs. journals
Journal-based ES > FDA-based ES for each of the 12 drugs Increase in effect size from FDA to Journal: - Minimum boost: 11% - Maximum boost: 69% - Average boost: 32%
FDA-focused investigations are limited in
their ability to provide crucial information
Investigators are discouraged from recruiting
samples that include a broad range of
depressive symptom severity
– Results cannot properly be generalized to the
population of depressed patients
– Informative tests of the relation of symptom severity to
the drug-placebo advantage cannot be conducted
Meta-Analyses of the FDA Database
(Khan et al. and Kirsch et al.)
Both showed that the drug-placebo difference substantial for
more severe patients, small at best for less severe cases
Most of the trials they included in their reports:
– Used a placebo washout design
– Excluded patients with HAMD of less than 20
Drug-placebo difference for typical depressed patient who is
given an antidepressant could not be estimated:
– In Khan et al., baseline mean < 23 in 0/45 studies
– In Kirsch et al., baseline mean < 23 in 2/27 studies
Fournier et al., JAMA 2010 Fournier, DeRubeis, et al. (2010, JAMA)
Drug-Placebo Differences
Clinical Significance (NICE)
Similar finding with psychotherapy (Driessen et al., 2013)
Efficacious: Elkin IPT; Dimidjian BA; Haringsma CDW; van Schaik IPT
Not Efficacious: Elkin CT; Dimidjian CT; Barlow SCT; Simpson DYN
FDA-focused investigations are limited in
their ability to provide crucial information
• Investigators are not encouraged to recruit samples that
represent the important subsets of depressed patients
– Results cannot properly be generalized to subsets of depressed
patients who are not included in studies
– Informative tests of the drug-placebo advantage in various
subsets cannot be conducted
Percentage of Patients who Responded
to 16 weeks of Treatment
58% 58%
0%
20%
40%
60%
80%
100%
ADM (120) CT (60)
Overall
Perc
en
t
Meds
Therapy
42%
Dropped out
or did not
respond
42%
Dropped out
or did not
respond
n = 65
n = 26
n = 29
n = 28n = 16 n = 15
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
0 1 2 + 0 1 2 +
ADM CT
Number of Prior Medication Attempts
Fin
al H
am
ilto
n S
co
rePrior Medication Treatment
predicts differential response in the two treatments
Meds
Ther.
Meds
Ther.
Employment Status predicts differential response in the two treatments
024
68
1012
1416
ADM (103) CT (47) ADM (17) CT (13)
Employed Unemployed
Fin
al H
am
ilto
n S
core
Diagnosis of Any Personality Disorder
ADM
CT
The diagnosis of any Personality Disorder predicts differential response in the two treatments
66%
44%49%
70%
0%
20%
40%
60%
80%
100%
ADM (59) CT (27) ADM (61) CT (33)
PD Non-PD
Re
sp
on
se
Ra
te
Fournier et al Brit J
Psychiatry 2008
67%
14%0%
20%
40%
60%
80%
100%
ADM CT
Clust B
Response R
ate
Patients Whose Personality Disorders Feature
Very Poor Emotion-Regulation Responded Well to
Medication, and Poorly to Cognitive Therapy
Percent of Patients who Remained Well
across Continuation
0%
20%
40%
60%
80%
100%
0 2 4 6 8 10 12
Time (months)
Su
rviv
al R
ate
0%
20%
40%
60%
80%
100%
0 2 4 6 8 10 12
Time (months)
Su
rviv
al R
ate
66%
44%
0%
20%
40%
60%
80%
100%
ADM (59) CT (27)
PD
Re
spo
nse
Ra
te
Response rates: Patients with Any Personality Disorder
ADM = 38% CT = 38%
Placebo = 6%
Percent of Patients who met Response Criteria
RELAPSE rates: Patients with Any Personality Disorder
FDA-focused investigations are limited in
their ability to provide crucial information
• Investigators are not encouraged to discern the effects of
the medicines, other than on depression symptom
severity and tolerability (side effects)
Change in Neuroticism
27
29
31
33
35
beginning 8 weekN
euro
ticis
m
placebo
Paxil
Change in depression severity
12
14
16
18
20
22
24
beginning 8 week
HR
SD
placebo
Paxil
Personality Change during Pharmacotherapy for
Depression: Drug-Placebo Differences
Tang et al Archives 2009
Personality Change during Pharmacotherapy for
Depression: Switch to Medications
12
14
16
18
20
22
24
26
beginning 8 week 16 week
HR
SD
27
28
29
30
31
32
33
34
35
36
beginning 8 week 16 week
Neu
ro
ticism
Depression Neuroticism
Placebo SSRIs
Tang et al Archives 2009
Placeb
o
SSRI
s
0.00
0.20
0.40
0.60
0.80
1.00
% o
f Re
spo
nd
ers
No
t Re
lap
sin
g
0 2 4 6 8 10 12
Time from End-of-treatment (Month)
––– top-third in neuroticism reduction
––– middle-third
––– bottom-third in neuroticism reduction
0.00
0.20
0.40
0.60
0.80
1.00%
of R
es
po
nde
rs N
ot R
ela
ps
ing
0 2 4 6 8 10 12
Time from End-of- treatment (Month)
0.00
0.20
0.40
0.60
0.80
1.00
% o
f Re
spo
nde
rs N
ot R
ela
ps
ing
2 4 6 8 10 12
Time from End-of- treatment (Month)
0.00
0.20
0.40
0.60
0.80
1.00
% o
f Re
spo
nde
rs N
ot R
ela
ps
ing
2 4 6 8 10 12
Time from End-of- treatment (Month)
Upper 1/3 of Δ
Neuroticism
(ns)
Lowest 1/3 of Δ
Neuroticism
(ns)
Middle 1/3 of Δ
Neuroticism
(p<.05)
Relapse during Continuation, by
Condition and Δ Neuroticism
––– medication continuation
––– continuation on placebo
Tang et al Archives 2009
To whom are antidepressants being
prescribed? What other options are there?
Most patients, especially in primary care, would
not qualify for pharmaceutical trials (Zimmerman et
al. J Clin Psychopharmacol, 2002) on basis of
severity
Look to the United Kingdom’s NICE guidelines for
mild depression: “Watchful waiting, guided self-
help, computerised CBT, exercise, brief
psychological interventions.”
Be skeptical of combination treatments, at least as
initial programs of care.
A Hot-Button Topic
Judith Warner: “The Wrong Story About
Depression” (NYTimes, 1/8/2010)
Richard Friedman: “…on close inspection, the new
study does not stand up to (the) mountain of
earlier evidence.” (NYTimes, 1/11/2010)
Peter Kramer: “In Defense of Antidepressants.” “In
the end, the much heralded overview analyses
look to be editorials with numbers attached.”
(NYTimes, 7/9/2011.)
Highly recommended
Acknowledgments
Support:
NIMH
GlaxoSmithKline
Collaborators:
Dan Strunk Jay Fournier
Tony Tang Lois Gelfand
Steve Hollon Jay Amsterdam
Lorenzo Luaces
top related